Pandora Information On The Pandora''''s Ext Port.


ashaw

Still Fresh
Joined
Jul 1, 2008
Messages
35
I was wondering what the voltage ratings on the uart, are ie what the IO voltages are.
Can it use 3.3 V IO.
 
I'd like to tack some questions on to this thread. (Don't worry if the replies are slow, ashaw. Hardware dudes are probably flat out.)

Ok, my EXT questions.

1. Will the stereo audio level be controlled by the volume wheel?

2. Is a 5V supply available at the EXT port? I need to power a small FM transmitter (for car kit). It would be nice to make up a custom EXT cable for both power and audio to the transmitter, but I'll take power from USB if I have to.
 
Last edited by a moderator:
'Gruso' said:
1. Will the stereo audio level be controlled by the volume wheel?
2. Is a 5V supply available at the EXT port?
1. No, it's a line level output right off the DAC
2. No, it's a 2.8V supply at up to 200mA


ashaw:
The I/O pins on the EXT port are all 1.8V logic. Level shifting is required to go higher. A simple to solder part like the TXB0104DR can be used to provide bidirectional transfers to a higher voltage part in your custom application.
 
Last edited by a moderator:
'MWeston' said:
2. No, it's a 2.8V supply at up to 200mA
ashaw:
The I/O pins on the EXT port are all 1.8V logic. Level shifting is required to go higher. A simple to solder part like the TXB0104DR can be used to provide bidirectional transfers to a higher voltage part in your custom application.


If the supply is 2.8V and the logic is 1.8V, will it cause problems to drive the TXB0104DR's VCCA @2.8V?
Eg. will 1 from the pandora be seen as 1 by the TXB? (it will see Vcca*0.65 as 1... which is 1.82v (eg. not sure?))
And will the pandora be damaged if its inputs are driven at 2.8V levels?

Would putting a diode in series with the TXB Vcca so that the TXB sees ~2.1V help?
- it would see 1.36v as 1 :)
- and its output to pandora would be max 2.1V
-which is just 0.3V above the level - usually not bad (?)
- (eg. remember atmels and some other things datasheets saying max input Vcc+0.3V etc)
 
Last edited by a moderator:
You've misunderstood the purpose of the TXB0104. It separates the two systems so that each one can connect together without burning out the other one or being too "low" to be read.

Here's the schematic for the breakout board: http://openpandora.ca/schematic/AV_Board_schematic.pdf

The TXB0104 is basically invisible to both sides of the communication and data can go both ways without any control signals and at up to 60Mbps.

I would imagine the OMAP3 chip has about a 10% tolerance above 1.8V, but doing stuff like that might shorten the life of the pin. People who want to hack on the port need to be smart about how they do things because this port is only protected against ESD and not DC voltages under about 6V that could burn out parts of the chip. Digikey sells this chip in single part quantities and they are quite cheap so I expect everyone to use one. :)
 
Last edited by a moderator:
'MWeston' said:
You've misunderstood the purpose of the TXB0104. It separates the two systems so that each one can connect together without burning out the other one or being too "low" to be read.
I think not. I was referring that you seemed to suggest 2.8v for the TXB0104's A side but the line level is 1.8v.
'MWeston' said:
Here's the schematic for the breakout board: http://openpandora.ca/schematic/AV_Board_schematic.pdf
Thanks for the breakout board schematics (i had looked for them but didnt find). :)

You use a LP2992 to regulate "VDD_IO_1V8" (1.8V) for the TXB0104s VCCA - this explains everything :)
What i was missing was essentially "How the 2.8v voltage is useful (directly to the TXB) if the logic is 1.8v?"
Another thing that keeps bugging be is that why 2.8V in the ext port? You already have 5V in the USB connector so why not 5V, it could also be regulated to 1.8v (for TXB0104) and give power to traditional (eg. ST232ACN) serial line level converters (and TXB0104's B side) (and µC's (though atleast atmels can live with 2.8v)).
Now I'm thinking that if i were to make my own "additional board" it will need
- a regulator for 1.8v (will anyways need this)
- a boost converter for 5V (I wouldn't neet this if ext power were 5V)
- my own 5V stuff (I can elaborate on this but i think that this isn't needed here.)
 
Last edited by a moderator:
Do you really need 5V because if you could use 3.3V i have a cuircuit for a boost buck all ready done up.
 
'ashaw' said:
Do you really need 5V because if you could use 3.3V i have a cuircuit for a boost buck all ready done up.
I was going to run an AVR (ATMega88) on the CTS/RTS lacking (AVR doesn't have them natively either) serial port to create a parallel port (needs 5V power) using a high serial speed of 1.8432 Mbps (which needs 14.7456Mhz for the atmel - which is above 10Mhz - which needs 4.5V-5.5V) (the OMAP3 can (from the OMAP35x tech manual) reach high serial speeds (max 3686400)) . I was also going to just add a normal ST232ACN (5V chip) serial level converter and a DE-9 (std serial) connector to the another (w flow control) serial port.
 
Last edited by a moderator:
Have you Seen the XMEGA A1 this can do 32Mhz at 3.3V, this is what i am intending to use.
 
STM32 here (www.st.com/stm32). My current favourite :) 72 MHz, although I've overclocked one to 100Mhz, and underclocked to 32Khz (Software programmable PLL). 2V - 3.6V.
 
Last edited by a moderator:
I was wondering, What connector does the EXT port have again?

Working on my project at the moment.
 
I think its this:CODE
http://catalog.tycoelectronics.com/catalog/bin/TE.Connect?S=24100&M=PPROP&P=&BML=10000-M&LG=1&PG=1&IDS=321950,321232,372004,372007,382493,321233,442539,437007&N=7
(1717169-2)
 
thankyou, would you like the files for my optimised voltage convertor, I am thinking about making a little board that contains a DC-DC boost convertor and a TXB0108 switchable between 3.3V and 5V the board is currently 21mm by 23mm, I think I can get this board to 2cm by 1cm if I put components on boath side of the board mabe even smaller, I will put a XMEGA A1 on this board too if it fits well.
 
'ashaw' said:
thankyou, would you like the files for my optimised voltage convertor, I am thinking about making a little board that contains a DC-DC boost convertor and a TXB0108 switchable between 3.3V and 5V the board is currently 21mm by 23mm, I think I can get this board to 2cm by 1cm if I put components on boath side of the board mabe even smaller, I will put a XMEGA A1 on this board too if it fits well.
I don't see a big need for the files now (for me), but ofcourse i could take a look at them... Anyways i have a few points to make here:
- I think you're making it all SMD - thats good for size but bad for my skills of soldering them* - so yeah i could be interested in boying this thing but fully assembled please, and then it could be a bit too expensive (dunno).
- And on the XMEGA - yeah it's fast and has lots of memory and peripherals, but it can only run upto 3.6V - so switching the board to 5V would fry it - not good. And as an another point it has 78 I/O's.. if you want to keep the size of the board to a minimum, you would need to skip routing a big majority of them, wich would negate a lot of the good features of the XMEGA, etc. I would suggest you put no microcontroller on the converter board and make a connector on the board to wich you can design your additional XMEGA board, and to wich other people can put their own favorite µC boards. (As a personal note: what atmel µC i would like to see on the board == minimum=tiny2313, good=mega[8,16,32]8, biggest=mega644)

* Mostly soldering the XMEGA A1 sounds very hard (think would need soldering paste and an oven) - the other components might go ok, but really i dont know, i dont have much experience on SMD soldering and no special tools for it...
 
Last edited by a moderator:
For BGA version, sure, but for TQFP version, it can easily be done with just a soldering iron and flux - I would just advise against trying to solder each pin individually. Many in-experienced people try and do that and it ends up taking them over an hour to solder it down, when it can be done in 5 minutes if you use flux (oh, and leaded solder. Don't use the unleaded stuff)

If you stick with 0805 components, it's not really that difficult. 0603 is where it becomes tricky.
 
Last edited by a moderator:
I have finished my board, Though i fear that the assembaly is not for us mere mortals it does have the proparty that it can be soldered into a dip header socket such as this one Mill Max 150-40-316-00-001000

Please find a pdf containing layout and a schematic at

http://www.alexisshaw.com/voltageconvertor.pdf

please give me some feedback as i intend to have a few of these made up and then to pot them, they are switchable by a rotary switch on the top, and have a built in boost convertorfor 3.3/5 V

PM me for gerbers and Drill files.
 
'ashaw' said:
I have finished my board, Though i fear that the assembaly is not for us mere mortals it does have the proparty that it can be soldered into a dip header socket such as this one Mill Max 150-40-316-00-001000

Please find a pdf containing layout and a schematic at

http://www.alexisshaw.com/voltageconvertor.pdf

please give me some feedback as i intend to have a few of these made up and then to pot them, they are switchable by a rotary switch on the top, and have a built in boost convertorfor 3.3/5 V

PM me for gerbers and Drill files.
Oh My God... That looks both a) Great! and B) truly not assemblable by me...
Now that the converter is in that format (installable to /in place of a DIL16 socket (did i get it right?)),
i can see other uses for it too.
Eg. it could be used for running 5V circuits from 3.3V (when set to 5V) with some I/O's, etc.
EDIT: Not without removing/bypassing the 1.8V requlator... (I just came to the forums index and then went :excl:)
And the MAX1675's datasheet looks very nice - 94% efficiency at 200mA (!!).
That means ~105 mA for the 5V circuit, (~159mA for 3.3V).

In short - looks great,
if they work, i would be willing to boy one or two of these (fully assembled) (depending on price).
 
Last edited by a moderator:
it looks like the price for these is about 20 US at qtys of 100 pcs getting a quote for the boars now. the 1.8 V converter can easily be replaced for the equivelant 3.3V part with only a chane in 1 part.

I will probably send it to an assembly house to have made up. this is not to be made up by mere mortals.

I need to have this looked on by someone that knows more about PCB design as this is my 1st PCB.
 
'ashaw' said:
it looks like the price for these is about 20 US at qtys of 100 pcs getting a quote for the boars now. the 1.8 V converter can easily be replaced for the equivelant 3.3V part with only a chane in 1 part.
I did take a second (longer) look at the pdf. Since you routed the 1.8V line out, it could be changed to handle boost&I/O from "any" single voltage by just desoldering (relatively simple thing i think when you know what you're doing...) the 1.8V regulator and then putting the single voltage to both UB and "1.8V" lines.
'ashaw' said:
I will probably send it to an assembly house to have made up. this is not to be made up by mere mortals.
I can see why...

'ashaw' said:
I need to have this looked on by someone that knows more about PCB design as this is my 1st PCB.
Truly WTF/Hard to believe. That PCB is fully SMD with 4 layers (if i read the pdf correctly - took a while to figure out why i couldn't "get it" (was trying to interpret as 2-layered)). Compare to: My first PCB was a single layered through-hole thing with very wide traces. It propably had a few 74-series chips and propably did something stupid (i don't really remember - it was a school work).

Anyways - i didnt find anything wrong with the schematics. Neither with the board but it was pretty hard to read for me (4 layers). Suggest you either trust yourself (not me) or wait until someone else takes a look on it (who knows more about PCB design).
 
Last edited by a moderator:
Back
Top